A familiar route to open-weight models
WeightsAPI brings model selection, API access and capacity planning together for published open-weight models. Managed inference puts model serving behind an API so an application can request an answer without operating the underlying machines. Open weights make a model available to inspect or deploy under its license; they do not mean every model has the same rights, behavior or capabilities. The catalogue brings those differences together so you can choose deliberately.
For people building useful applications
Independent builders, development teams and product teams often need the same first answers: which model fits the task, what a request may cost, and how it will fit an existing application. WeightsAPI organizes that starting point around a playground, model pages and practical examples. Whether you are exploring an assistant, a coding workflow or document processing, begin with a small representative example and clear success criteria before committing more time or traffic.
Two ways to plan inference capacity
Shared inference uses input and output token rates and suits variable workloads. The dedicated offering provides monthly ordering for a defined workload, subject to confirmed hardware availability and deployment scope. Protect credentials, decide what data to send, check generated results and control the actions taken from them. A managed endpoint has finite capacity. Compare the two using your workload, model compatibility and pricing information.
Make the important details visible
Exact model identifiers, visible input and output rates, and documented limits help you assess each option. Model pages distinguish publisher specifications from service limits that require deployment confirmation. Check availability separately from the catalogue, and validate the behavior your application needs. The trust page explains data handling and account controls.
What you can use today
Browse the catalogue, compare published prices, read the documentation and manage API keys after signing in. Place a dedicated GPU order, pay from confirmed account credit and follow its status until provisioning is confirmed. Crypto checkout provides payment instructions; credit requires transaction verification. Check service status for model availability and the FAQ for account and payment steps.
Know who to contact before committing
WeightsAPI is the service name. The support page provides troubleshooting guidance and a report you can copy; copying does not submit a request. Official operator and contact details have not been published. Confirm the operator, applicable terms and data-handling arrangements before making a service commitment.